You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "David Capwell (Jira)" <ji...@apache.org> on 2024/05/01 21:57:00 UTC
[jira] [Commented] (CASSANDRA-19182) IR may leak SSTables with pending repair when coming from streaming
[ https://issues.apache.org/jira/browse/CASSANDRA-19182?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17842786#comment-17842786 ]
David Capwell commented on CASSANDRA-19182:
-------------------------------------------
test results are not clean but tried to track down if this existed before my branch or not. Some are due to the CI environment (3 test classes use the wrong methods so JAVA_TOOL_OPTIONS break them) but others are failing and known already...
> IR may leak SSTables with pending repair when coming from streaming
> -------------------------------------------------------------------
>
> Key: CASSANDRA-19182
> URL: https://issues.apache.org/jira/browse/CASSANDRA-19182
> Project: Cassandra
> Issue Type: Bug
> Components: Consistency/Repair
> Reporter: David Capwell
> Assignee: David Capwell
> Priority: Normal
> Fix For: 4.0.13, 4.1.5, 5.0-beta2
>
> Attachments: ci_summary-cassandra-4.0-258e3327d63438a2f7b836bd044547649f7eeaf6.html, ci_summary-cassandra-4.1-a3d85e9995654b202895e64d3cbd23a7caf57655.html, ci_summary-cassandra-5.0-a618e90ec8a7a0b125b69aff5c25081a17b10bc4.html, ci_summary-trunk-13afd4a6936c9441f96979d3e9b50afdeda80ea1.html, result_details-cassandra-4.0-258e3327d63438a2f7b836bd044547649f7eeaf6.tar.gz, result_details-cassandra-4.1-a3d85e9995654b202895e64d3cbd23a7caf57655.tar.gz, result_details-cassandra-5.0-a618e90ec8a7a0b125b69aff5c25081a17b10bc4.tar.gz, result_details-trunk-13afd4a6936c9441f96979d3e9b50afdeda80ea1.tar.gz
>
> Time Spent: 1h 40m
> Remaining Estimate: 0h
>
> There is a race condition where SSTables from streaming may race with pending repair cleanup in compaction causing us to cleanup the pending repair state in compaction while the SSTables are being added to it; this leads to IR failing in the future when those files get selected for repair.
> This problem was hard to track down as the in-memory state was wiped, so we don’t have any details. To better aid these types of investigation we should make sure the repair vtables get updated when IR session failures are submitted
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@cassandra.apache.org
For additional commands, e-mail: commits-help@cassandra.apache.org